INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CONTRACTION-INDUCED SKELETAL muscle injury is associated with stretch of highly activated muscle (lengthening contractions; 1, 19, 23, 37). Fortunately, training with lengthening contractions protects muscle from injury during subsequent lengthening contractions (7, 13, 24, 26, 33-35). Despite the consensus for the existence of exercise-induced protection from muscle injury and its potential importance, little is known about this phenomenon.

Armstrong and colleagues (1, 35) proposed that a population of injury susceptible fibers undergoes necrosis following lengthening contractions. Training with lengthening contractions may eliminate susceptible fibers, or parts of fibers, leaving muscle fibers that are less susceptible to injury. In addition, Devor and Faulkner (9) reported that muscles injected with bupivacaine to induce degeneration or regeneration showed dramatic protection from contraction-induced injury. This observation is consistent with the hypothesis that degeneration or regeneration is sufficient to protect muscle from contraction-induced injury. Whether degeneration or regeneration is necessary to induce protection has not been established.

There is little evidence of whether training that does not involve lengthening contractions nor measurable injury can protect muscle from lengthening contraction-induced injury. A single bout of training with maximal voluntary isometric contractions protected muscle from enzyme leakage during a subsequent identical bout of isometric exercise (6). Whether maximal isometric training can protect muscle from injury following lengthening contractions is not known. Previous exposure to level running that did not appear to produce injury in vastus intermedius and triceps brachii muscles protected these muscles from injury during downhill running in rats (35). Although the authors speculated that the lengthening contractions during level running produced the protective effect, isolating the effects of a specific type of contraction using the treadmill running model is not possible.

The hypothesis of this study was that lengthening contractions and subsequent muscle fiber degeneration and/or regeneration are required to induce exercise-associated protection from lengthening contraction-induced muscle injury. A corollary to this hypothesis was that training with isometric contractions or stretches of passive muscle that did not produce fiber degeneration and/or regeneration would not induce protection.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals. Three- to-four-month-old specific pathogen-free male C57BL/6 mice (Harlan Sprague-Dawley, Indianapolis, IN) were housed in a pathogen-free barrier facility until experimentation and then returned to a separate barrier facility between experimental procedures. All experimental procedures were approved by the University Committee for the Use and Care of Animals at the University of Michigan.

In situ pretraining evaluation. Each mouse was anesthetized with an intraperitoneal injection of 2% avertin (0.015 ml/g body wt). Supplemental doses (0.1 ml) were administered if the mouse responded to a toe pinch. A small incision was made at the right ankle, and the distal tendons of the extensor digitorum longus (EDL) muscle were exposed. The mouse was then placed on a heated platform maintained at 37°C. The hindlimb was stabilized by fixing the distal femur between sharpened screws and securing the foot with tape to the platform. The intact tendon was tied with 4-0 silk to the lever arm of a servomotor (Aurora Scientific, Richmond Hill, ON, Canada), which controlled the length of the muscle and measured the force developed by the muscle. The servomotor was controlled by a computer to move the lever arm through a given distance at a constant velocity. The computer was also used to collect and analyze force data. The small region of exposed tendon was kept moist with warmed isotonic saline.

Training bout and functional evaluation. The EDL muscle was exposed to one of three training protocols: stretches with the muscle near maximally activated (lengthening contractions), stretches with the muscle relaxed (passive stretches), or near-maximal activation of the muscle held at L_o (isometric contractions). Near-maximal activation was achieved using the voltage determined in In situ pretraining evaluation to elicit a maximal response and a frequency (150 Hz) that elicited 94 ± 1% P_o. A sham training group underwent identical surgical procedures, but no stretch or contractions were performed. All training protocols involved 75 repetitions performed at 0.25 Hz for a total exercise duration of 5 min. Lengthening contractions and passive stretches were initiated at L_o and were of 20% strain relative to L_f, at a velocity of 1 L_f/s. For lengthening contractions, stretch was initiated 100 ms after the beginning of stimulation from the plateau of force development. Ten minutes after the 75 repetitions, P_o was remeasured. The incision at the ankle was closed with 7-0 nylon and bathed with povidone-iodine solution, and then mice were allowed to recover from anesthesia. For some mice in each group, mice were anesthetized 3 days after the training bout and isometric contractile properties were measured in situ using a protocol identical to that for the pretraining evaluation. These mice were then killed by cervical dislocation immediately after functional testing, and EDL muscles were removed for histological evaluation (see Histological evaluation). These mice provided data for nontrained muscle that could be compared with those for trained muscle. Force deficits were calculated for each mouse as the decrease in maximum isometric force produced 10 min and 3 days following the training bout expressed as a percentage of pretraining values.

Posttraining injury bout and functional evaluation. Two or six weeks after the training bout of lengthening contractions, passive stretches, or isometric contractions, EDL muscles were exposed in situ to 75 lengthening contractions using a protocol identical to the training bout of lengthening contractions. Three days after the posttraining injury bout, mice were anesthetized, isometric contractile properties were measured in situ using a protocol identical to that for the pretraining evaluation, and force deficits were calculated in the same manner as those for the training bout.

Histological evaluation. After the final in situ functional evaluations, anesthetized mice were killed by cervical dislocation. EDL muscles were dissected, blotted dry, weighed, mounted in tissue freezing media, and frozen in isopentane chilled with dry ice. Cross sections were cut at 10 µm with a cryostat and stained with hematoxylin and eosin. Fibers exhibiting clear evidence of degeneration or regeneration and the total number of fibers in a single cross section from the mid-belly of each muscle were counted with the aid of a microscope imaging system (Bioquant, Nashville, TN). Degenerating fibers were defined as those with infiltration of inflammatory cells, pale and/or discontinuous staining of the cytoplasm, or a substantially swollen appearance. Regenerating fibers were defined as those with nonperipheral nuclei without other degenerative changes. Fiber counts were performed by two independent observers, and values were averaged. Degenerating and regenerating fibers are reported as a percentage of the total number of fibers in the cross section.

Data analysis. Physiological cross-sectional area (PCSA) was calculated by dividing muscle mass by the product of L_f and muscle density, 1.06 mg/mm³. Specific P_o was calculated by dividing P_o by physiological cross-sectional area. Mean values and standard errors were determined for each variable for each group. Force deficits and percentages of injured fibers were compared between groups using the nonparametric Kruskal-Wallis one-way ANOVA on ranks (multiple groups) or the Mann-Whitney ranked-sum test (two groups). The 0.05 level was taken to indicate statistical significance. A significantly smaller force deficit in trained vs. nontrained muscle indicated protection from functional injury, and a significantly smaller percentage of injured fibers indicated protection from structural injury.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

For 51 mice included in the study, mean body mass (27.3 ± 0.4 g) and noninjured EDL muscle mass (10.2 ± 0.1 mg), L_o (12.6 ± 0.1 mm), P_o (379 ± 7 mN), and specific P_o (22.7 ± 0.3 N/cm²) were similar to those reported previously for mice of the same strain and age (4). Ten minutes after the training protocols, lengthening contractions resulted in a force deficit of 59.0 ± 6.1%, passive stretches resulted in a much smaller force deficit of 9.8 ± 1.4%, and isometric contractions resulted in a force deficit not significantly different from zero. The absence of a force deficit after isometric contractions suggests that the force deficit associated with lengthening contractions was not due to surgical procedures or to metabolic fatigue. The force deficits measured 10 min after the training protocols involving stretch of active or passive muscle may have been influenced by a shift in L_o to longer muscle lengths (15), because P_o was measured at the L_o determined before the protocol for both the pre- and 10 min posttraining evaluations. The force deficits at 3 days were not influenced by any shift in L_o, because L_o was determined anew at the beginning of the evaluation at 3 days, and P_o was measured at this newly determined L_o. In fact, 3 days after training with lengthening contractions, L_o was 12.9 ± 0.1 mm, ~2% greater than before training. Two weeks after training, however, L_o was not different from the pretraining value.

Three days after the training protocol, lengthening contractions resulted in a force deficit of 55.1 ± 5.0%, whereas passive stretches and isometric contractions resulted in force deficits not different from zero (Fig. 1A). Compared with cross sections from nonexercised control EDL muscles, cross sections from EDL muscles 3 days after lengthening contractions contained a greater percentage of injured fibers (0.8 ± 0.2 vs. 18.3 ± 2.9%; Fig. 1A). Passive stretches and isometric contractions produced percentages of injured fibers not different from control. Some muscles that had undergone training with passive stretches did show some evidence of enlarged blood vessels and edema between muscle fibers (Fig. 1B). There were very few (<0.3%) regenerating fibers in any of these groups.

View larger version (75K): [in this window] [in a new window]   Fig. 1.   A: force deficits and percentages of injured fibers in cross sections of extensor digitorum longus (EDL) muscles 3 days after bouts of 75 passive stretches (n = 6), isometric contractions (n = 6), or lengthening contractions (n = 8). Force deficit calculated as percent decrease in maximum isometric force produced before and after bout. Percentage of injured fibers calculated as percentage of the total number of fibers in a cross section determined to exhibit clear signs of degeneration. Values are means with SE bars. * Significant difference from zero; ** significant difference from control (P < 0.05). B: cross sections of EDL muscles stained with hematoxylin and eosin 3 days after no exercise, passive stretches, isometric contractions, and lengthening contractions. Note substantial change in morphology of muscle following lengthening contractions and enlarged blood vessel (v) in muscle following passive stretch. A muscle spindle is also present (s).

Two weeks after the training bout of lengthening contractions, muscles had fully recovered force production as values for P_o were not different from preinjury values (the force deficit, 5.2 ± 4.3%, was not significantly different from zero). The second bout of lengthening contractions performed 2 wk after the training bout produced a force deficit and a percentage of injured fibers at 3 days that was only one-third of values for untrained muscles (Fig. 2). Six weeks after the training bout, the second bout of lengthening contractions produced a force deficit and a percentage of injured fibers at 3 days that were not different from values for untrained muscle (Fig. 2). The percentages of regenerating fibers 2 and 6 wk after the initial bout were 15.1 ± 5.2 and 8.0 ± 1.7%, respectively.

View larger version (64K): [in this window] [in a new window]   Fig. 2.   A: force deficits and percentages of injured fibers in cross sections of EDL muscles 3 days after an injury bout of 75 lengthening contractions in groups that were not trained (n = 8), 2 wk posttraining (n = 7), or 6 wk posttraining (n = 7), with an identical bout of lengthening contractions. Force deficit calculated as percent decrease in maximum isometric force produced before and after bout. Percentage of injured fibers calculated as percentage of total number of fibers in a cross section determined to exhibit clear signs of degeneration. Values are means with SE bars. Force deficits and percentages of injured fibers for all groups were significantly greater than zero (P < 0.05). * Significant difference from nontrained group (P < 0.05). B: cross sections of EDL muscles stained with hematoxylin and eosin 3 days after an injury bout of 75 lengthening contractions for mice that were not trained, 2 wk posttraining, or 6 wk posttraining with an identical bout of lengthening contractions. Section representing nontrained group is same as that used for lengthening contraction group of Fig. 1.

Two weeks after the training bouts of either isometric contractions or passive stretches, the bout of lengthening contractions produced a force deficit at 3 days that was two-thirds of the values for untrained muscles (Fig. 3). For muscles trained with passive stretches, the percentage of injured fibers at 3 days was one-half of that for untrained muscles (Fig. 3). For muscles trained with isometric contractions, the percentage of injured fibers at 3 days was not significantly different from values for untrained muscle (Fig. 3). There was a significantly greater percentage of injured fibers and a trend of a larger force deficit (P = 0.06), for muscles trained with passive stretches or isometric contractions compared with those trained with lengthening contractions. The percentage of regenerating fibers was very small 2 wk after either passive or isometric training (<0.3%). Two weeks after sham training, a bout of lengthening contractions produced a force deficit (54.7 ± 7.5%) and a percentage of injured fibers (18.1 ± 3.1%) at 3 days that was not different from those for untrained muscle, suggesting that the surgical procedures did not induce protection from injury.

View larger version (69K): [in this window] [in a new window]   Fig. 3.   A: force deficits and percentages of injured fibers for EDL muscles 3 days after an injury bout of 75 lengthening contractions in groups that were not trained (n = 8), 2 wk posttraining with passive stretches (n = 7), or 2 wk posttraining with isometric contractions (n = 10). Force deficit calculated as percent decrease in maximum isometric force produced before and after bout. Percentage of injured fibers calculated as percentage of total number of fibers in a cross section determined to exhibit clear signs of degeneration. Values are means with SE bars. Force deficits and percentages of injured fibers for all groups were significantly greater than zero (P < 0.05). * Significant difference from nontrained group (P < 0.05). B: cross sections from EDL muscles stained with hematoxylin and eosin 3 days after an injury bout of 75 lengthening contractions for mice that were not trained, 2 wk posttraining with passive stretches, or 2 wk posttraining with isometric contractions. Section representing nontrained group is same as that used for lengthening contraction group of Fig. 1.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The major finding of this study was that neither lengthening contractions nor muscle fiber degeneration and/or regeneration was necessary to induce exercise-associated protection from lengthening contraction-induced injury. Training with isometric contractions and passive stretch protected muscle from injury, as evaluated by the force deficit at 3 days for both training protocols and by the percentage of injured fibers at 3 days for the passive stretch protocol. That the isometric contraction or passive stretch protocols did not produce muscle fiber degeneration and/or regeneration is supported by three sets of observations. First, there was no significant percentage of injured fibers in muscle cross sections 3 days after the protocols. Second, there was also no significant force deficit at 3 days. Third, there was no significant percentage of regenerating fibers in muscle cross sections 2 wk after these protocols. Cross sections of some muscles subjected to passive stretches did show evidence of enlarged blood vessels and edema at 3 days, suggesting that some damage may have occurred, perhaps to blood vessels or to connective tissue. Whether such potential damage or resulting edema plays a role in inducing protection is a question for future investigation. Although training with isometric contractions or passive stretches provided protection, training with lengthening contractions tended to provide greater protection. Lengthening contractions or degeneration and regeneration may thus produce stronger stimuli for protection than isometric contractions or passive stretches.

A previous study reported that in vivo endurance isometric training (10 Hz stimulation, 1 s on/1 s off duty cycle, 30 min/day, 5 days/wk, 3 wk) did not protect rabbit dorsiflexor muscles from lengthening contraction-induced injury as evaluated by the force deficit (31). Comparing these previous results with those from the present study suggests that isometric exercise involving relatively small forces may not provide a sufficient stimulus to induce protection; high force isometric contractions may be necessary. Further investigation is required to determine whether there is a force threshold at which isometric contractions will provide protection from lengthening contraction-induced injury.

The magnitude and time course of protection induced by lengthening contractions are in agreement with those reported previously for the mouse tibialis anterior muscle (34). In the present study, the protective effect of a bout of lengthening contractions, indicated by the reduction of the force deficit by two-thirds, was evident at 2 wk but not detected at 6 wk. Sacco and Jones (34) reported that a bout of 120 in vivo lengthening contractions produced a force deficit of 48% (measured in vitro) 3 days after the bout. Three weeks after the initial bout, a second bout of lengthening contractions produced a force deficit of only 20% 3 days after the second bout. By 12 wk, a second bout produced a force deficit at 3 days not significantly different from that for the initial bout. Whether the protective effect of the initial bout was lost earlier than 12 wk was not determined in the previous study. The magnitude of protection induced by lengthening contractions in the present study was also consistent with that produced by bupivacaine injection (9), suggesting that degeneration or regeneration may induce similar protective mechanisms in the two models.

One proposed mechanism of exercise-induced protection from the primary injury is an increase in the number of sarcomeres in series in muscle fibers (20, 42). Downhill run training has been reported to increase sarcomere number in the rat vastus intermedius muscle (20). An increase in sarcomere number would 1) decrease sarcomere extension for a given muscle lengthening and 2) shorten sarcomere length for any given muscle length and thus move sarcomeres away from the descending limb of the force-length relationship. Morgan (25) has argued that the descending limb is inherently unstable and that because of this instability, muscle is particularly susceptible to injury when lengthening contractions are performed at long lengths (e.g., 27). However, experiments using training with controlled lengthening contractions have revealed no significant increase in sarcomere number in rabbit dorsiflexor muscles (16) or in rat EDL muscles (Brooks, unpublished data). In the present study, there was no significant increase in L_o 2 wk after any of our training protocols. In addition, for the injury protocol, lengthening contractions were initiated at L_o, and stretch parameters were scaled to L_o (magnitude, velocity). Consequently, lengthening contractions were likely performed over approximately the same portion of the force-length relationship for each group, and any potential increase in sarcomere number in this study would not account for the protection observed.

Injury after lengthening contractions takes place in two stages: a primary insult directly associated with the contractions and delayed, secondary damage that peaks ~3 days after the exercise (12, 18, 26). The primary injury appears to be mechanical in nature (2, 5, 40), perhaps a result of inhomogeneities in the strengths of sarcomeres in series, resulting in strong sarcomeres pulling weak sarcomeres apart (21, 25, 28). A potential mechanism for exercise-induced protection from the primary injury is increasing the strength of the cytoskeletal protein network that surrounds sarcomeres (39) and transmits tension through the membrane (30, 38). Upregulation of proteins in this network (e.g., desmin, talin, vinculin, dystrophin) could help to stabilize sarcomeres during lengthening contractions and thus protect muscle fibers from injury. For example, lengthening contractions resulted in increased talin and vinculin concentration in rat soleus muscles (14). Talin and vinculin are also upregulated, along with utrophin, another cytoskeletal protein, in regenerated muscle of mdx mice (17). Whether upregulation of these proteins is directly involved in protecting muscle from contraction-induced injury has yet to be determined.

The secondary injury may be related to the inflammatory response and free radical damage (10, 36). Following lengthening contractions, inflammatory cell invasion has been associated with increased reactive oxygen species generation (3) and increased oxidation of glutathione (22). A potential mechanism for protecting muscle from the secondary injury is upregulation of free radical scavenging pathways. Administration of exogenous superoxide dismutase reduced the force deficit and percentage of injured fibers following lengthening contractions in the EDL muscles of young and old mice (43), suggesting that superoxide or its byproducts contribute to the injury process. Although 12 wk of high intensity level treadmill running increased both superoxide dismutase and glutathione peroxidase activity in rat soleus muscles (8), regulation of antioxidant activity following lengthening contractions or passive stretches has not been explored.

Parts of the initial and delayed force deficits may be a result of mechanically induced or free-radical-induced damage to the excitation-coupling apparatus and resulting failure (29, 41). Thus exercise-induced protection from lengthening contraction-induced force deficit may be partly due to protection from excitation-coupling failure (24). Whether training protects muscle from excitation-coupling failure remains to be determined.

In conclusion, we found that neither lengthening contractions nor muscle fiber degeneration and/or regeneration were necessary to induce exercise-associated protection from lengthening contraction-induced injury. However, lengthening contractions and associated degeneration or regeneration did produce greater protection than training with passive stretch or isometric contractions. The mechanisms underlying the protection provided by passive stretch and isometric training, and those associated with the greater protection induced by lengthening contractions, await further investigation.

Perspectives